The following article, “A Comprehensive New Data Analysis Into Why Harris Lost in 2024” by Amy Walter, is cross-posted from The Cook Political Report:

After every major presidential and midterm election, the Democratic data firm Catalist releases a comprehensive analysis of the composition and partisan leanings of the electorate. What distinguishes their analysis from election night exit polling is that it integrates data like vote history, Census data and Catalist’s own polling and modeling, which can give us a more fulsome view of who joined the electorate, and — as importantly — who dropped out of it.

The Catalist “What Happened” report, shared first exclusively with The Cook Political Report, finds that Vice President Kamala Harris’ 2024 loss was due to two key factors: She lost roughly two points of support among those who turned out in 2020, and new and infrequent voters failed to materialize for Democrats as they had in previous elections.‬ For the first time in Catalist’s dataset, both infrequent and new voters — groups that tend to be younger and more diverse than the electorate at large — fell below 50% support for a Democratic presidential candidate.

What Happened

The Obama Coalition Turned Into the Trump Coalition

Back in 2012, Barack Obama’s campaign had a mantra — a younger, more diverse electorate was the key to reelection. Had the 2024 electorate been in place in 2012, the team in Chicago would’ve been over the moon. The electorate last fall was three points less white than it was 12 years ago. What’s more, the youngest cohort of voters by generational breakdown made up 36% of the electorate (compared to 18% in 2012), while the oldest cohort was just 39% (down from 57% in 2012).

And, yet, that coalition was much friendlier to Donald Trump than to any other GOP nominee.

So what happened?

Men — Especially Men of Color — Shifted Towards Republicans

Overall, Harris performed six points worse among men than Biden did. But that falloff was significantly more pronounced among Latino men (-12) and Black men (-7). Among white men without a college degree — a group that is traditionally the least friendly to Democrats — the slippage was least dramatic (-3).

Women Didn’t Rally for Harris, and Latina Women Moved Right

Despite the historic nature of her candidacy as the first woman of color to be nominated for president, women greeted that fact with more of a shrug. Harris’ vote share among women was basically the same as the share Biden got in 2020, except for a noticeable decline in support from Latinas. She also slightly underperformed Biden among Asian American and Pacific Islander (AAPI) women (-4) and failed to do any better with Black women than Biden did.

Younger Voters, Especially Voters of Color, Shifted Right

Even as many in the TikTok generation embraced Harris’ “coconut tree” meme, she failed to capture the imagination or support from young voters that Biden enjoyed in 2020. Among all young voters, Harris underperformed Biden by six points, but the gap was much larger among younger Latino (-12) and AAPI (-9) voters. Among Black and white younger voters, the drop-off in support from 2020 was less severe (-4).

How it happened

For much of the 2024 election, the Harris campaign was convinced that they could win if the electorate looked similar to 2020, while the Trump campaign was eager to expand the electorate. In other words, Harris wanted an electorate heavily populated with frequent voters, while the Trump team wanted an electorate filled with voters who have not participated as much in major elections.

According to Catalist, the electorate was indeed more populated with frequent voters than at any point in the last three presidential elections. The share of the electorate that were so-called “super voters” — those who voted in all of the last four major elections — was 47%, compared to just 38% of the electorate in 2020. More importantly, Harris did better among these voters than any previous Democratic nominee since 2016, capturing 50% of the vote, compared to Biden’s 49% and Hillary Clinton’s 47%.

The least frequent voters also made up less of the electorate overall in 2024 (11%) than in 2020 (16%) and 2016 (15%).

Both of those data points — the large share of “super voters” and the smaller share of less frequent voters — suggest that Harris should have been successful in 2024.

So why wasn’t she?  Because Harris not only underperformed Biden among those who were brand new to voting, but also among people who had voted in anywhere from one to three of the last four elections.

Super Voters Were Older, Less Racially Diverse, and Less Urban

So why did Harris underperform with all types of infrequent voters, not just those who hadn’t voted in any previous election? The Catalist data shows that less frequent voters are more likely to be younger, live in an urban area, be a person of color, and not hold a college degree. In other words, there is a direct correlation between Harris’ underperformance with younger, urban and voters of color and the fact that they make up a disproportionate share of infrequent voters.

For example, in 2024, 78% of the voters who voted in the last four elections were white, compared to 67% of those who voted in two of the last four elections and just 62% who voted in none of the last four elections.

A Lot of Biden Voters Stayed Home, and New Voters Were the Least Democratic Ever

For the last 12 years, Catalist analysis has found dropoff voters (those who voted in the previous presidential election but don’t show up in the current one) have been disproportionately Democratic-leaning. But in the last three presidential elections, Democrats made up for that loss by replacing them with brand new voters, who also leaned heavily Democratic. This is how Democrats were able to win the popular vote in every one of those elections.

In 2024, however, that “churn” in the electorate didn’t benefit Harris. According to Catalist, 30 million 2020 voters didn’t cast a ballot in 2024. That 30 million was also a very Democratic-leaning group, giving Biden an estimated 55.7% of the vote in 2020. Harris was unable to make up for those lost votes with new voters because 1) there were only 26 million new voters; four million fewer than those who dropped out and 2) the new voters gave Harris just 48.5% of the vote, falling below 50% for a Democratic presidential candidate for the first time in Catalist’s dataset.

It’s these “churn” voters that Catalist concludes both parties should be focused on in 2026 and beyond.

“The Republican playbook has historically counted less on refreshing these rotating marginal voters,” writes Catalist, “but these data suggest that may become a more important part of their electoral coalition.”

And, while they suggest that “future Republican candidates may not be able to replicate Trump’s performance,” future Republican campaigns now have a roadmap to victory — something they didn’t have before Trump re-imagined the GOP coalition.

Note: Catalist is a Democratic organization, but its data and analysis is trusted across partisan lines and among election data experts. Their partisanship does mean that their data is delivered entirely in terms of Democratic vote share rather than a comparison of the two parties. For more on Catalist’s methodology, click here.